Strength of in-situ wood 4

[JAE]

We have an old brick and wood framed building where we would like to estimate the floor load carrying capacity.

In our review, we have documented the span, size and spacing of wood floor joists as well as applied dead loads. What we don't know is the Fb of the wood.

Is there an ASTM test procedure that we could give to a testing lab to test some samples of the wood? Or any other procedure in the industry?

We thought we might do a two-point bending test on each piece, mapping the force/deflection curve to determine an Fu and then use some safety factor (2?) to determine Fb.

 

[boo1]

JAE,
ASTM D143, D245, D1990, 2555, and D4761 are the appliable standards.

ASTM D143 is one of two basic mechanical testing Standards for wood in North America when the goal is to develop estimates of properties for structural design. ASTM D143 tests small clear (knot- and defect-free) specimens then modifies the clearwood properties and their variability to estimate the 5th percentile of clearwood property then further account for the effects of knots, slope-of-grain, density, and grain orientation (flat-, quarter-, or bastard-sawn). With flat-sawn lumber the growth rings are parallel to widest face. With quarter-sawn lumber the growth rings are perpendicular to widest face. With bastard-sawn material there is NO control on growth ring orientation. Bastard sawn most accurately reflects actual production lumber. ASTM D245 instructs the user how to use D143 values to derive allowable design stress (ADS) values. ASTM D4761 is used when testing production-sized lumber and timber. ASTM D1990 instructs the user how to use D4761 values to derive ADS values. Both methods are used to derive allowable design stress values in North America. Most hardwoods use the D143/D245 methods, most softwoods use the D4761/D1990 method. The former was used from 1934-1990, while the latter is now the preferred way.

Consider a 3-point flexural test method.

see the proceedures listed at:

http://www.fpl.fs.fed.us/DOCUMNTS/PDF1997/green97a.pdf

http://www.fpl.fs.fed.us/documnts/FPLGTR/fplgtr113/Ch06.pdf

 

[JAE]

boo1 for president!

 

[boo1]

I thought you would like the FPL files, as the addressesed your issues.....

 

[RockEngineer]

Boo1 is master once again. I am always amazed at your knowledge of where to find things on the web and the vast knowledge you have at your finger tips.

 

[boo1]

I love that Forest Products Laboratory site and the "Wood handbook--Wood as an engineering material", it's like a $150 text book.

 

[boo1]

see

http://www.wwpa.org/pdf/TN4.pdf

http://www.wwpa.org/pdf/TN5.pdf

 

[Tmoose]

I think I'd get a wood person to determine the wood species and grade the beams, then use table values.

Depending on species (Visual) grading can whack the allowable fb over 50%. In my 1930 Mark' Handbook some of the grading criteria are density, rings-per-inch, size of knots, size of timber, checks (visible), and shake (can be invisible damage from the tree hitting the ground). They also mention the sample-to-sample variability is likely to be >20%.

If I tested an unchecked clear specimen I guess I would expect to get an allowable fb that was much higher than any published spec, or any real beam.

The few amateur wood floor system calcs I have done hit deflection limits rather than strength limits.

Alos related to deflection, If you have machinery or lots of foot traffic there will be resonant frequency/vibration issues that will require good planning to avoid.

 

[boo1]

Mr Cohen recommended Wood Advisory Services (see

http://www.woodadvisory.com/),

contact Al DeBonis. Although they are located in Millbrook, NY, the sample can be shipped.

 

[whyun]

boo gets my star too.

JAE, in lieu of testing, have you looked at the code that was in effect when the building was designed? You may be able to find the values you need tabulated.

 

[RARWOOD]

I think I would follow the approach TMOOSE suggested with a few minor modifications.

I would establish an estimated grade based on the current grade rules for the species of wood that is in place. I would then use the current N.D.S. design values for that grade a species applying a 0.9 factor to account for long term loading.

The reason I would follow the above procedure is because the strength of the wood is determined by the strength reducing characterisics such as knots, splits decay pockets ect, which are present in the wood.

Many people say that lumber today is not as strong as it was in the past. This is based on a belief that "Old Growth Timber" is better than second growth timber.I am not aware of any research which supports this belief.

In fact the In-grade testing program, which was at its time, and still maybe, the largest testing program conduct to evaluate the strength of the wood, produced results contrary to the above believe.

However returning to the question of establishing design values, I would follow the above outlined procedure just to see how close to an estimated stress limit the existing joist were. In many cases wood members are oversized for architectural reasons. So if I was well below the estimated design limit I would not worry about it.

However the above suggestions for additional testing are all good ideas.

Sometimes though an in place load test to a level acceptable to the local Building Official, is the best solution.

 

[SacreBleu]

RARSWC,
Knots are actually a limb growing out of the trunk. Some years ago, the allowable stress values for 2x lumber were drastically reduced according to the "new growth" concept, which mainly said that harvested trees were much smaller in trunk diameter. The 2x lumber would have defects that went thru the entire thickness, much more so than 4x and 6x sawn pieces.
The allowable bending stresses were slightly "tweaked" in the 1991 and 1997 NDS Supplements for Mechanical Properties. In the 2001 NDS, the allowable shear stresses were dramatically increased, perhaps in recognition of new research about wood failing by horizontal shear.

 

[SlideRuleEra]

2001 NDS allowable shear strength was increased because a factor for splits, checks and/or shakes was mistakenly being applied twice. With the "removal" of one factor, allowable shear "appears" to double. See page two on this link

http://www.awc.org/Publications/papers/NDS2001article.pdf

http://www.SlideRuleEra.net